MINERVA MEDICA COPYRIGHT® REVIEW
Some current issues in the pharmacokinetics/pharmacodynamics of antimicrobials in intensive care N. PETROSILLO 1, C. M. DRAPEAU 1, M. AGRAFIOTIS 2, M. E. FALAGAS
2, 3, 4
1“L. Spallanzani”, National Institute for Infectious Diseases Rome, Italy; 2Alfa Institute of Biomedical Sciences (AIBS), Athens, Greece; 3Department of Medicine, Henry Dunant Hospital, Athens, Greece; 4Department of Medicine, Tufts University School of Medicine, Boston, MA, USA
ABSTRACT Infections, particularly those caused by resistant pathogens, are a common cause of morbidity and mortality in critically ill patients. However, the availability of effective antimicrobial agents is limited. Critical illness itself can influence the pharmacokinetic/pharmacodynamic (PK/PD) parameters of antimicrobials by altering their volume of distribution and the rate of their excretion and elimination and by impairing their penetration into tissues. Therefore, when designing a treatment regimen, the intensivist should consider and take advantage of antibiotic PK/PD properties. There is significant but inconclusive evidence that critically ill patients may benefit more when antibiotics with time-dependent action are administered in a continuous/prolonged infusion regimen. On the other hand, aminoglycosides exhibit a concentration-dependent pattern of killing and should be administered at high doses once daily or at extended intervals, and their levels in the plasma should by strictly monitored to avoid both underexposure and toxicity. The problem of antimicrobial resistance now involves agents traditionally considered reliable in that aspect, such as vancomycin. Strict monitoring of vancomycin MIC for methicillin-resistant Staphylococcus aureus and the prudent use of the available alternative agents as well as de-escalation strategies might be reasonable strategies for dealing with this problem. (Minerva Anestesiol 2010;76:509-24) Key words: Pharmacokinetics - Anti-infective agents - Intensive care.
I
nfections occurring in intensive care units (ICU) are associated with high morbidity and mortality rates.1-4 This can be partly attributed to the fact that critically ill patients are highly susceptible to infections due to co-morbidities and the impairment of mechanical and immunological protective barriers.5 Indeed, critically ill patients could be compared to an unbalanced system because of their variable and instable clinical status, such that they require prompt and continuous therapeutic adjustments to maintain organ function, homeostasis and clinical stability. In addition, ICU facilities themselves have become a réservoir of aggres-
Vol. 76 - No. 7
sive and multi-resistant pathogens, while the availability of effective antibiotics is often very limited, a fact related mainly to the extensive and inappropriate use of antimicrobial agents. Methicillinresistant Staphylococcus aureus (MRSA),6 and multi- or pan-drug-resistant Pseudomonas aeruginosa, Acinetobacter baumannii and Enterobacteriaceae 7 are characteristic examples of aggressive pathogens that the ICU physician must confront in everyday clinical practice. Moreover, the inappropriate use of antibiotics, together with the severity of the underlying conditions of critically ill patients, is responsible for poor clinical outcome.8
MINERVA ANESTESIOLOGICA
509
MINERVA MEDICA COPYRIGHT® PETROSILLO
CURRENT ISSUES IN ANTIMICROBIALS IN INTENSIVE CARE
Given these factors, optimizing antibiotic efficacy in the ICU setting has become mandatory. This target can be reached by taking best advantage of the pharmacokinetic/pharmacodynamic (PK/PD) properties of antibiotics. Indeed, it has been demonstrated that PK/PD properties are major determinants of the in vivo efficacy of antibiotics, and data from in vitro and animal studies have demonstrated that incorrect manipulation of PK/PD parameters may lead to the emergence of resistance.9 These findings have also been corroborated in the clinical setting. For example, Thomas et al. have shown that an AUC0-24/MIC ratio less than 100 was strongly associated with the selection of antimicrobial resistance in a cohort of acutely ill patients treated with ciprofloxacin and various beta-lactams.10 Over the years, the literature has provided extensive information about the PK/PD properties of single antibiotics; however, new PK/PD issues that have recently emerged render antibiotic therapy even more complex and controversial, especially in the ICU setting.11 In this article, some current issues concerning the PK/PD properties of antibiotics commonly used for the treatment of infections in the ICU setting will be discussed, with a major emphasis on the route of administration and the minimum inhibitory concentration (MIC) values. It is true that a few more potent antimicrobial agents (e.g., linezolid and tigecycline) have become available for the treatment of severe ICU infections during the last years. However, these innovations do not compensate for limited availability of effective antimicrobial agents. Due to the emergence of multi-drug resistance pathogens, clinicians are likely to switch to new antimicrobial agents; however, the problem of antimicrobial resistance is still evolving. The emergence of multi-drug resistant pathogens can be averted or delayed if clinicians take into account the particular PK/PD characteristics of the commonly administered agents and consider pathogen susceptibility patterns.8-10 For these reasons, this review will focus on the PK/PD aspects of commonly used antimicrobials in the ICU setting, such as beta-lactams, vancomycin and aminoglycosides, and will provide particular examples on how the proper administration of older agents and the inclusion of newer ones in de-escalation strategies
510
might become a powerful tool in dealing with the problem of emerging antimicrobial resistance. Prior to discussing PK/PD issues related to antimicrobial treatment of the critically ill, a brief introduction to basic PK/PD principles of antimicrobial chemotherapy will be provided. Basic PK/PD principles of antimicrobial chemotherapy Pharmacokinetics aims to quantify the time course of the serum level of an agent by employing, among others, parameters such as the drug peak serum (Cmax), steady-state serum (Css) trough serum level (Cmin), volume of drug distribution (Vd) and the area under the serum concentrationtime curve (AUC). On the other hand, pharmacodynamics quantifies the activity of an antimicrobial agent by integrating its PK parameters with the MIC for a particular pathogen. From the PD point of view, antibiotics can be categorized based on their mode of bacterial killing and the presence of a postantibiotic effect. Thus, the pattern of bacterial killing of an antibiotic can be concentration-dependent if higher concentrations of the agent result in more extensive elimination of the pathogen or timedependent, if the effectiveness of bacterial killing depends upon the duration of pathogen exposure to the agent. The term “post-antibiotic effect” (PAE) refers to the time required by the pathogen to resume normal growth following exposure to the agent.12, 13 Therefore, antimicrobial agents can be classified basically into two categories. The first category includes drugs that exhibit concentration-dependent killing in combination with a prolonged postantibiotic effect (e.g., aminoglycosides, fluoroquinolones); the best predictors of efficacy for this class of agents are the peak drug concentration divided by the MIC (Cmax/MIC, e.g., aminoglycosides) 14, 15 and/ or the AUC at 24 hours in relation to MIC (AUC0-24/MIC, e.g., fluoroquinolones).16, 17 The second category includes drugs that exhibit a time-dependent pattern of killing with minimal or moderate post-antibiotic effect (e.g., beta-lactams, macrolides, glycopeptides); for this category, the time the concentration of antibiotic remains above the MIC (T>MIC, e.g., beta-lactams) 18 and/or AUC0-24/MIC (e.g., glycopeptides) 19, 20 are the parameters most strongly correlated to clinical
MINERVA ANESTESIOLOGICA
July 2010
MINERVA MEDICA COPYRIGHT® CURRENT ISSUES IN ANTIMICROBIALS IN INTENSIVE CARE
PETROSILLO
TABLE I.—Principal pharmacodynamic/pharmacokinetics characteristics of antimicrobials A.
B.
According to the pattern of antimicrobial killing: 1.
Time-dependent antibiotics (Beta-lactams, glycopeptides, linezolid, quinupristin/dalfopristin and the glycylcyclines).The time that free antimicrobial concentrations remain above the MIC (T > MIC) is the PK/PD index correlating with efficacy. PAE is minimal (Beta-lactams) or moderate (glycopeptides, linezolid, quinupristin/dalfopristin, glycylciclines).
2.
Concentration-dependent antibiotics (aminoglycosides, fluoroquinolones, daptomycin. The peak concentration/minimum inhibitory concentration (Cmax/MIC) ratio and/or the area under the concentration–time curve at 24 h/MIC (AUC0–24/MIC) ratio are the best PK/PD parameters correlating with efficacy. Moreover, there is a prolonged postantibiotic effect (PAE).
According to solubility: 1.
Hydrophilic antibiotics (Beta-lactams including penicillins, monobactams, cephalosporins and penems, glycopeptides, aminoglycosides): their volume of distribution is limited to the extracellular space and their plasma and interstitial concentrations may decrease due to fluid extravasation. They are inactive against intracellular germs, and have renal elimination.
2.
Lipophilic antibiotics (Macrolides, fluoroquinolones, tetracyclines and glycilcicline, oxazolodinones, rifampicin, cloramphenicol): they have a large volume of distribution, and the dilution of interstitial fluids is less relevant compared to hydrophilic antibiotics. They are active against intracellular germs and are mainly eliminated by the liver.
efficacy (Table IA). Furthermore, antibiotics can be classified as either hydrophilic or lipophilic based on their ability to cross cellular membranes and the resultant volumes of distribution (Table IIB).12, 13 Influence of critical illness on PK/PD parameters of antibiotics Critical illness is characterized by significant alterations in 1) fluid distribution and homeostasis; 2) hemodynamic parameters and microcirculation; and 3) organ function scores. These abnormalities may affect PK/PD parameters in a variety of unpredictable ways. Antibiotic distribution volume Fluid distribution and homeostasis undergo significant derangements in critically ill patients via three main pathophysiologic processes: 1) fluid extravasation, as in sepsis, trauma, hypoalbuminemia, external fluid overload, renal and cardiac failure; 2) fluid loss, as in surgical drainages and burns; and 3) local fluid overload, as in pleural infusion and ascites. These disturbances result in an increased Vd for hydrophilic antibiotics with a subsequent decrease in their plasma concentration.21 Lipophilic drugs, on the other hand, are not significantly influenced by these alterations due to their large volumes of distribution.13
Vol. 76 - No. 7
Antibiotic excretion and elimination The cardiac index may be normal or even increased in severe sepsis and septic shock, particularly following fluid resuscitation.13 Thus, unless organ dysfunction ensues, renal artery blood flow is also increased, resulting in the enhanced delivery and excretion of hydrophilic and moderately lipophilic antibiotics,11 decreasing their half-life. In addition, hypoalbuminemia, which is frequently encountered in critically ill patients, may further augment antibiotic clearance by increasing the free fraction of protein-bound antibiotics (e.g., teicoplanin, ceftriaxone).13 Lastly, the administration of drugs with hemodynamic modes of action (such as dopamine or dobutamine) or diuretics may also influence the glomerular filtration rate and consequently alter antibiotic clearance.13 On the other hand, advanced critical illness is characterized by multiple organ failure, with the kidneys and the liver commonly involved in this process. Acute kidney damage is very common in critically ill patients 22 and may prolong the elimination half-life of the renally-excreted hydrophilic and moderately lipophilic drugs, leading to an accumulation of toxic metabolites. Furthermore, continuous or intermittent renal replacement therapy, a measure commonly utilized in ICUs for the management of renal failure, can significantly alter antibiotic clearance via a variety of mechanisms related to the mode of treatment, exchange rates,
MINERVA ANESTESIOLOGICA
511
MINERVA MEDICA COPYRIGHT® PETROSILLO
CURRENT ISSUES IN ANTIMICROBIALS IN INTENSIVE CARE
membrane properties and duration of therapy.23 On the other hand, the elimination of lipophilic antibiotics is mainly influenced by liver dysfunction. However, the liver has a considerable functional reserve and dose adjustments are not usually required except in more severe cases of liver failure.24 However, the impact of the liver support systems on antimicrobial PK/PD parameters remains undefined.25 Tissue penetration Septic shock may significantly affect antibiotic distribution to the tissues by decreasing their concentrations at the target site to subinhibitory levels even while the achieved plasma drug concentrations would still be considered “effective”.26 This situation, though still largely unexplored, could account for clinical treatment failures. Continuous/prolonged or intermittent infusion of antimicrobials for critically ill patients in the ICU setting? Continuous/prolonged administration of antibacterial agents with a time-dependent pattern of killing is considered a plausible and enticing therapeutic option. Kasiakou et al.27 have evaluated PK/PD data provided from 17 randomized controlled trials (most of them involving critically ill patients) that compared continuous and intermittent modes of administration of antibacterials with time-dependent action (beta-lactams or vancomycin). They found that the Css achieved with a continuous infusion strategy was higher than the Cmax achieved via intermittent infusion in all 14 studies that provided relevant data; furthermore, the T>MIC was higher with continuous infusion in 3 out of 6 studies that assessed this issue. They concluded that the continuous infusion strategy might be a better option for infections caused by bacteria with high MIC values, a problem relevant to the ICU setting.27 In addition, the same group has evaluated clinical outcomes in a meta-analysis of randomized controlled trials that compared continuous and intermittent infusions of various antimicrobial agents (beta-lactams, aminoglycosides and vancomycin).28 This study showed a tendency towards a better clinical out-
512
come for the continuous infusion arm (OR 0.73, 95% CI 0.53-1.01) that reached statistical significance when studies which used same total daily doses in both arms were analyzed separately (OR 0.70, 95% CI 0.50-0.98; P=0.004); in addition, no differences in mortality or nephrotoxicity were noted.28 This issue was revisited recently by Roberts et al. in a meta-analysis involving 14 RCTs investigating continuous or prolonged versus intermittent administration of beta-lactam antibiotics.29 This study failed to find any statistically significant difference in clinical cure or mortality rates in favor of the prolonged or continuous administration strategy. However, this result can also be attributed to differences in total doses between the two arms, a failure to reach clinically relevant PK/PD therapeutic targets in individual studies and the inclusion of a diverse population of patients.29 Indeed, critically ill patients might represent a separate subgroup that could particularly benefit from prolonged or continuous infusion of antibacterial agents with time-dependent modes of action.30 Some issues concerning antibiotics that exhibit a time-dependent pattern of killing and are commonly used in the ICUs will be analyzed further. Vancomycin Vancomycin has played a major role in the treatment of bacterial infections caused by multidrugresistant gram-positive pathogens, most notably methicillin-resistant Staphylococcus aureus (MRSA).31, 32 Vancomycin exhibits a time-dependent mode of bacterial killing with a moderate postantibiotic effect. Knudsen et al. demonstrated that T>MIC and Cmax are the parameters that may predict the clinical efficacy of a single-dose glycopeptide treatment.19 Furthermore, a seminal paper by Moise-Broder et al. emphasized the importance of vancomycin AUC0-24/MIC in predicting clinical success (>350) and bacterial eradication (>400).20 However, vancomycin penetrates poorly into the tissue, and concerns have been raised regarding its clinical efficacy against deep-seated infections.33, 34 For instance, the “Tarragona strategy” strongly discourages the use of vancomycin for the treatment of ventilator-associated pneumonia (VAP) caused by MRSA and other gram-positive pathogens.35 Therefore, some authors propose that
MINERVA ANESTESIOLOGICA
July 2010
MINERVA MEDICA COPYRIGHT® CURRENT ISSUES IN ANTIMICROBIALS IN INTENSIVE CARE
severe infections caused by MRSA should be managed with a rapid achievement and maintenance of trough (minimum) plasma concentrations (Cmin) of 15 to 20 mg/L.36, 37 Commonly used vancomycin administration schedules often fail to achieve a Cmin of more than 15 mg/L.38, 39 Indeed, Kitzis et al. have demonstrated that the administration of vancomycin in two or four daily doses often failed to achieve a Cmin of more than 15 mg/L.38 In this study, Cmin was 3 hour infusion) meropenem infusion. They demonstrated that, regarding Enterobacteriaceae, the probability of attaining bacteriostatic (30% T>MIC) and bactericidal (50% T>MIC) exposures were high for both intermittent (0.5, 1 and 2 g every 8 hours) and prolonged (0.5, 1 and 2 g every 8 hours; 1 and 2 g every 12 hours) infusion regimens. Thus, for the same dose of meropenem (1 g every 8 hours), prolonging the infusion time from 30 minutes to 3 hours would increase bactericidal activity against Enterobacteriaceae only slightly, if at all (0% change for E. coli, Enterobacter cloacae and Serratia species and 0.5% change for Klebsiella
MINERVA ANESTESIOLOGICA
July 2010
MINERVA MEDICA COPYRIGHT® CURRENT ISSUES IN ANTIMICROBIALS IN INTENSIVE CARE
pneumoniae). However, against Acinetobacter and Pseudomonas species, the highest target attainment rates were achieved with a prolonged infusion regimen at a dosage of 2 g every 8 hours.54 The same issue was also addressed in a 2 500subject Monte Carlo simulation by Lomaestro and Drusano.55 In their model, three dosing regimens, a 1-hour infusion of 0.5 g of imipenem/cilastatin delivered every 6 hours, a 3-hour infusion of 0.5 g meropenem delivered every 8 hours and a 3hour infusion of 1 g meropenem delivered every 8 hours were tested against strains of methicillinsusceptible S. aureus (MSSA), Klebsiella, Serratia, Enterobacter, Acinetobacter species and P. aeruginosa with MIC values for imipenem/cilastatin and meropenem ranging from 0.25 to 32 mg/L. All three dosing regimens were equally successful in effecting high rates of bactericidal exposure against MSSA and Enterobacteriaceae. However, against Acinetobacter species and P. aeruginosa, optimal bactericidal exposure was achieved with a 3-hour infusion of 1 g meropenem delivered every 8 hours.55 Furthermore, Kotapati et al.56 compared the clinical and economic outcomes of two meropenem intermittent (MICs for MICs of 1, 4, 8, and 16 mg/L were calculated from the individually fitted concentration-time curves. This study showed that optimal bactericidal exposure against pathogens with an intermediate resistance (MIC of 16 mg/L) could only be achieved with a 3-hour infusion of 2000 mg of meropenem delivered every 8 hours.57 Piperacillin/tazobactam (P/T) is another betalactam antimicrobial widely used in the ICU that could also be considered for continuous or prolonged infusion. Notably, P/T solutions have previously been shown to be stable at 37°C for at least 24 hours.58 Tam et al.59 performed a retrospective analysis in order to evaluate the effectiveness of P/T monotherapy in cases of bacteremia due to P. aeruginosa with P/T MIC values at the highest end of susceptibility as defined by the Clinical Laboratory Standards Institute (32 mg/L and 64 mg/L). They conducted two parallel studies, the first of which included cases of bacteremia due to P. aeruginosa strains with MIC values 32 or 64 mg/L (N=7) and the second of which focused on isolates with MIC values 15, a longer hospital stay before the first positive culture and P/T treatment were all factors significantly associated with a higher risk for 30-day mortality in the logistic regression analysis. Despite their disappointing findings, the authors asserted that P/T might be a reasonable treatment option for P. aeruginosa bacteremia, although alternative dosing regimens such as prolonged or continuous infusion might be considered in treating strains with high P/T MIC values.59 However, data on the clinical efficacy of piperacillin administered by continuous infusion are scant.60-64 In a study involving 40 septic critically ill patients, Rafati et al.60 demonstrated that, when the MIC was 16 mg/L, the target %T>MIC was achieved in all patients treated with continuous infusion (2 g over 0.5 h as a loading dose, then
MINERVA ANESTESIOLOGICA
515
MINERVA MEDICA COPYRIGHT® PETROSILLO
CURRENT ISSUES IN ANTIMICROBIALS IN INTENSIVE CARE
8 g daily over 24 h) but only in 62% of those patients treated with intermittent infusion (3 g every 6 hours over 0.5 hours). Moreover, piperacillin continuous infusion was associated with a more rapid resolution of infection compared to intermittent infusion as demonstrated by APACHE II scores (P=0.04).60 In another study, Bulitta et al.64 reported PK/PD data for three different modes of piperacillin administration: short infusion (3 g in 30 min every 4 h), prolonged infusion (3 g in 4 hours every 8 hours) and continuous infusion (8 g in 24 hours) in 8 adult patients with cystic fibrosis. In the Monte Carlo simulation, they found that %T>M was 100% at MIC≤16 mg/L for continuous infusion, while for intermittent infusion it varied from 55% to 100% depending on MIC values.64 A controlled study by Grant et al.61 involving 98 hospitalized patients with community or hospital-acquired infections reported on the clinical success rates of P/T continuous and intermittent infusion strategies. They found a non-significant higher success rate for the continuous infusion mode versus the intermittent one (94% vs. 82%; P=0.081), as well as a significantly lower number of days to fever resolution and lower costs for the continuous infusion group.61 Other studies, however, produced contrasting results. Li et al.62 randomized patients with complicated intra-abdominal infections caused by pathogens with a low a piperacillin/tazobactam MIC to receive either P/T by continuous infusion (13.5 g over 24 h) or intermittent infusion (3.375 g every 6 h). Based on data from 56 patients for whom the pharmacokinetic analysis was available, they concluded that the infusion method had no influence on PK parameters.62 A randomized study on 262 hospitalized patients with complicated intra-abdominal infections was performed by Lau et al.63 in order to compare piperacillin-tazobactam continuous infusion (12 g/1.5 g administered continuously over 24 h) with the standard intermittent infusion strategy (3 g/0.375 g administered over 30 min every 6 h). After analyzing 167 clinically evaluable patients, they noted that the rates of clinical cure or improvement at the test-of-cure visit for patients treated with continuous infusion or intermittent infusion were 86.4% and 88.4%, respectively (P=0.817).
516
Additionally, bacteriological success and defervescence were similar between the two groups.63 Recently, Boselli et al.65 randomized 40 patients with microbiologically documented VAP to receive P/T in a continuous infusion at a daily dose of 12/1.5 g or 16/2 g. Serum and alveolar concentrations of piperacillin/tazobactam were recorded for the two groups. The authors reported that the percentage of epithelial lining fluid (ELF) penetration for the two dosing regimens was 40–50% and 65–85%, respectively. Both serum and alveolar P/T concentrations were correlated to the patients’ creatinine clearance, suggesting wide pharmacokinetic variability. They also noted that continuous P/T infusion at a dose of 12/1.5 g/day might lead to insufficient antibiotic ELF concentrations in patients with no/mild renal insufficiency suffering from VAP caused by pathogens with high P/T MIC values; they suggested that, in that case, a continuous P/T dose of at least 16/2 g/day might be more appropriate. In patients with moderate/advanced renal failure, both dosages (16/2 g and 12/1.5 g) achieved serum concentrations far above the 35-40 mg/L MIC threshold.65 Recently, Lorente et al.66 performed a retrospective study of 83 patients with VAP caused by gramnegative bacteria who received initial empirical antibiotic therapy with P/T. Thirty-seven patients were treated with continuous infusion while 46 were treated with intermittent infusion; no difference existed between the two groups in terms of demographic characteristics, severity scores, comorbidities or renal function. A significantly higher rate of clinical cure was noted for the patients treated with continuous infusion compared to those treated with intermittent infusion (33/37 [89.2%] vs. 26/46 [56.5%]; P=0.001). However, there were no significant differences in mortality rate and in the duration of mechanical ventilation or in the length of ICU stay. Notably, logistic regression analysis showed that the probability of VAP clinical cure was higher with the continuous infusion strategy when the culprit pathogen had a MIC at the higher value of susceptibility, i.e., 8 mg/L (P=0.049) or 16 mg/L (P=0.03).66 Piperacillin penetration into the tissue of critically ill patients with sepsis was evaluated in a prospective randomized study of 13 critically ill adult patients with known or suspected sepsis.67
MINERVA ANESTESIOLOGICA
July 2010
MINERVA MEDICA COPYRIGHT® CURRENT ISSUES IN ANTIMICROBIALS IN INTENSIVE CARE
Patients were randomized to receive piperacillin/ tazobactam either via bolus administration (4 g/0.5 g every 6 or 8 hours) or via continuous administration (day 1: 4 g/0.5 g piperacillin–tazobactam bolus infusion followed immediately by a continuous 24-hour infusion of 8 g piperacillin/1 g tazobactam; day 2 and onward: 12 g/1.5 g piperacillin-tazobactam administered by 24-hour infusion). Patients treated with continuous administration had statistically significantly higher median plasma concentrations on day 2 compared to patients treated with bolus administration (16.6 vs. 4.9 mg/L; P=0.007), even though they were given a 25% lower piperacillin dose. Median tissue concentrations were not statistically different on day 1 and day 2 between the two groups. The authors concluded that the tissue pharmacodynamic targets were achieved more successfully with the continuous infusion regimen.67 The efficacy of an extended infusion dosing strategy for P/T in treating P. aeruginosa infections was explored in a retrospective study by Lodise et al.68 The authors employed Monte Carlo simulation to identify the optimal method of administering P/T for the treatment of P. aeruginosa infections. Their analysis revealed that the probability of attaining a 50% T>MIC (near bactericidal effect) was higher when P/T was administered in a 4-hour infusion at dose of 3.375 g every 8 hours compared to a 30-min infusion of 3.375 g every 4 to 6 hours. These two dosing regimens were validated in a cohort of 194 patients (65% of which were ICU patients), where 102 patients received the extended infusion regimen. All strains were susceptible to P/T, according to the Clinical Laboratory Standards Institute criteria (i.e., MIC≤64 mg/L). Classification and regression tree analysis revealed that an APACHE II score ≥17 was the most significant predictor of 14-day mortality. Overall, no statistically significant difference was observed between the two groups in terms of 14-day mortality and median length of hospital stay. However, when patients were stratified according to the severity of their illness, a significantly lower 14-day mortality (P=0.04) and median length of hospital stay (P=0.02) was evidenced for patients with APACHE II ≥17 who received the prolonged infusion regimen. The authors concluded that a P/T prolonged infusion regimen should be favored for the treatment of P. aeruginosa
Vol. 76 - No. 7
PETROSILLO
infections, particularly in critically ill patients.68 This study also highlighted the notion that critically ill patients are the most likely to benefit from strategies that optimize drug exposure and pharmacodynamics. Therefore, the data favoring continuous/prolonged administration of antimicrobials with timedependent patterns of killing in the intensive care unit are strong though not conclusive. Further studies should probably focus on critically ill patients with severe infections, a group most likely to benefit from this dosing regimen.30 Once daily (or extended interval) aminoglycoside regimens for critically ill patients Aminoglycosides constitute one of the oldest classes of antimicrobial agents. Their activity follows a concentration-dependent pattern with a significant post-antibiotic effect.69 In the 1980s, Moore et al. advanced the use of Cmax/MIC as a marker of aminoglycoside efficacy; they proposed that a Cmax/MIC value of 10-12 is associated with a higher probability of treatment success.14, 15 Following this work, Kashuba et al. have shown that attainment of a Cmax/MIC ≥10 results in a faster clinical improvement of patients with pneumonia caused by gram-negative pathogens.70, 71 In addition, this strategy was shown to be associated with a lower probability of selecting resistant strains.72 These results strongly supported the rationale behind the once daily (or extended interval) dosing schedule for aminoglycosides. Several studies have suggested that this dosing regimen is as effective as intermittent dosing and carries a lower risk of toxicity.73 A weight-based nomogram using a once-daily dose of 7 mg/kg for gentamycin and tobramycin and allowing dose interval adjustments on the basis on of a single subsequent plasma level assessment (obtained 6-14 hours following the start of aminoglycoside infusion) has been proposed by Nicolau et al.74 This dosing strategy was designed to maximize the efficacy of aminoglycosides against Pseudomonas aeruginosa and aims to achieve a peak concentration of 20 mg/L at one hour following drug administration, assuming a P. aeruginosa MIC40 mL/min/1.73 m2 who were administered gentamycin or tobramycin once daily at a dose of 7 mg/kg and found that the “Hartford” nomogram correctly predicted the dosing interval in all but one patient (98%). In addition, the mean Vd for their patients was 0.28±0.09 L/kg, similar to the value reported in the study by Nicoalu et al.74 They concluded that the “Hartford” nomogram is applicable to this subset of patients.80 Barletta et al.81 evaluated PK/PD parameters in a cohort of critically ill trauma patients treated with gentamycin or tobramycin once daily. The mean Vd of their patients was 0.3L/kg, with an intersubject variability of 33.8%. Four out of the 19 studied subjects experienced prolonged drugfree intervals (>12 hours). The investigators cautioned against the use of the “Hartford” nomogram in critically ill trauma patients and suggested individualized dosing based on at least two serum drug concentrations.81 Furthermore, the PK/PD parameters of gentamycin and tobramycin given once daily at a dose of 7 mg/kg were investigated in a prospective study by Bujik et al.82 on medical and surgical ICU patients. As expected, they observed lower clearances and higher drug half-life times for patients with a reduced creatinine clearance (
